Figure 1.
The original capsule network
| Procedure | Routing algorithm |
|---|---|
| 1: | procedure ROUTING |
| 2: | for all capsule i in layer l and capsule j in layer (l + 1):bij ←0 |
| 3: | for r iterations do |
| 4: | for all capsule i in layer l : ci ← softmax (bi ) |
| 5: | for all capsule j in layer (l +1) : sj ← |
| 6: | for all capsule j in layer (l +1) : vj ← squash (sj ) |
| 7: | for all capsule i in layer l and capsule j in layer (l + l) |
| 8: | return vj |
Figure 2.
CBAM Module
Figure 3.
Channel Attention Module
Figure 4.
Spatial Atttion Module
Figure 5.
Capsule network based on attention mechanism
TABLE I.
NETWORK MODEL AND PARAMETERS
| Layer | Parameters |
|---|---|
| Conv | inputChannel:1; outPutChannel:256 kernel size= 9; stride=1 |
| CBAM | Channel:256 |
| PrimaryCaps | InputChannel:256; OutputCaps:32*6*6 output_dim:8;kernel_size:9,stride:2 |
| MgitCaps | Inputeaps:32*6*6; out_put_caps:10 |
TABLE II.
NETWORK MODEL ACCURACY
| Net Work Name | Routing Number | CBAM Module | Max Accuracy | First time |
|---|---|---|---|---|
| CapsNetl | 1 | False | 99.70999908% | 133 |
| CapsNetl_CBAM | 1 | True | 99.66999817% | 100 |
| CapsNet2 | 2 | False | 99.65000153% | 111 |
| CapsNet2_CBAM | 2 | True | 99.61000061% | 42 |
| CapsNet3 | 3 | False | 99.69000244% | 131 |
| CapsNet3_CBAM | 3 | True | 99.62002324% | 82 |
| CapsNet4 | 4 | False | 99.59999847% | 141 |
| CapsNet4_CBAM | 4 | True | 99.55999756% | 86 |
| CapsNet5 | 5 | False | 99.65000153% | 125 |